Predictive modeling of software quality for very large telecommunications systems

Abstract

Society's reliance on large complex telecommunications systems mandates high reliability. Controlling faults in software requires that one can predict problems early enough to take preventive action. Software metrics are the basis for such predictions, and thus, many organization are collecting volumes of software metric data. Collecting software metrics is not enough. One must translate measurements into predictions. This study systematically presents a methodology for developing models that predict software quality factors. The individual details of this methodology may be familiar, but the whole modeling process must be integrated to produce successful predictions of software quality. We use two example studies to illustrate each step. One predicted the number of faults to be discovered in each module, and the other predicted whether each module would be considered fault-prone. The examples were based on the same data set, consisting of a sample from a very large telecommunications system. The sample of modules represented about 1.3 million lines of code.